This action might not be possible to undo. Are you sure you want to continue?
HOW INSURERS HAVE SHIFTED RISK AND COSTS ASSOCIATED WITH WEATHER CATASTROPHES TO CONSUMERS AND TAXPAYERS
J. Robert Hunter, Director of Insurance Consumer Federation of America
February 16, 2012
Insured losses from catastrophes around the globe totaled an estimated $108 billion in 2011, the second highest year in history. More than $30 billion of those losses occurred in the United States, likely the fifth or sixth most expensive year on record. Since 2004, storms like Katrina, Rita, Wilma and Ike, combined with other events have resulted in nearly $200 billion in catastrophe claims paid to millions of home, business and vehicle owners. Robert Hartwig, President of the Insurance Information Institute1
The question of how insurers deal with weather catastrophes, especially in years in which multiple events occur, has serious policy implications for Americans. In short, how can insurers handle all this risk, and is it legitimate to shift these costs to consumers and taxpayers? While insurance executives frequently remind the public and regulators of the frequency and severity of catastrophic events, industry data demonstrates that insurers have significantly and methodically decreased their financial responsibility for these events in recent years and shifted much of this risk to consumers and taxpayers. Some of the savings they have achieved is the result of the use of reinsurance and wise risk diversification strategies. However, most these savings have been achieved by hollowing out the coverage in homeowners insurance policies and raising rates. Insurers have also exposed taxpayers to more disaster assistance payouts and shifted high risk homes to state pools. This study investigates and analyzes the significant weather catastrophe risk-shift that has occurred in the last twenty years and offers recommendations to stop insurers from continuing to illegitimately shift costs and risks to taxpayers and consumers. CATASTROPHES: ONCE A SERIOUS PROBLEM FOR INSURERS The fact is that catastrophic weather events were once a serious problem for insurers. Consider the following charts:2
P/C INSURANCE ECONOMIC CYCLE
25.0% Op. Income as % of Premium 20.0% 15.0% 10.0% 5.0% 0.0% 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009
―III Response to Americans for Insurance Reform Report,‖ December 15, 2011. The data underlying these charts can be viewed at Addendum A.
PURE NET LOSS RATIOS
120.0% 100.0% 80.0% 60.0% 40.0% 20.0% 0.0% 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 P/C L/R HO L/R
―P/C L/R‖ is total property/casualty insurance losses divided by premium, called the ―loss ratio.‖ ―HO L/R‖ is the loss ratio for the homeowners’ insurance line of insurance.
The first chart illustrates the insurance industry economic cycle, showing operating income as a percentage of premium for the entire property-casualty insurance business over the last thirty years. There is a strong cyclical pattern to the industry’s results. Periodically, insurers’ profits decline to the break-even point. This is followed by what is known as a ―hard market,‖ in which coverage is hard to get and prices rise sharply. For example, a hard market began in 1975. Profits rose quickly thereafter and then, slowly, declined during the soft market until 1985, when another hard market started as profits dropped to zero and even a bit below that. A soft market began in 1987 and stayed in place until profits bottomed out again in 2001. The market is still soft as 2012 begins, but declining profits indicate that a hard market might be on the way. In fact, insurers are hoping for a hard market soon.3 One noteworthy aspect to the first chart is the sharp drop in overall property-casualty profits in 1992. What caused that one-year deviation from the normal cycle? The answer is that Hurricane Andrew adversely affected the insurance industry. Overall property-casualty profits fell that year by seven points as a direct result of Andrew. This is exactly what one would expect when a huge catastrophe occurs, because this is why Americans buy insurance, to cushion such occasional blows. The impact of Hurricane Andrew can also be clearly seen in the second chart. Net loss ratios of the property-casualty industry increased by about seven points because homeowner’s insurance profits were reduced by a whopping 40 points by Andrew. These charts demonstrate that at one time -- when hurricane Andrew hit in 1992 -insurers bore much of the financial risk of hurricanes. This trend clearly changed in the last decade, in which seven of the most destructive ten disasters in American history occurred, according to the Insurance Information Institute.4. The huge hurricane damages of 2004 (four
―Repeat Offenders: How the Insurance Industry Manufactures Crises and Harms America,‖ Americans for Insurance Reform, December 2011. 4 See http://www.iii.org/facts_statistics/catastrophes-us.html.
Florida hurricanes) and 2005 (Katrina and other hurricanes) had almost no impact on the overall property-casualty loss ratio or even the homeowners insurance loss ratio, as shown by the above chart. One factor that illustrates the trend of large events having minimal impact on insurers is the increasing surplus that property casualty insurers have accumulated in recent years. One would expect that, in years when large hurricane events occurred, insurers losses would increase, leveling out or even decreasing surplus over time. This has not occurred.
700 600 500 400 300 200 100 0 YEAR 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
INSURERS HAVE NOW “MASTERED” CATASTROPHIC EVENTS When four hurricanes hit Florida in 2004 and Hurricane Katrina pummeled the Gulf Coast in 2005, there was no noticeable impact on the overall profits or loss-ratios of propertycasualty insurers in either year. An examination of just the loss ratios for homeowners’ insurance in those two years shows an impact from the storms that is not noticeable. According to the Insurance Information Institute,5 Hurricane Andrew resulted in overall losses of $28 million, of which $17 million (64 percent) were paid out in insured losses. Hurricane Katrina resulted in overall losses of $125 million, of which insurers paid out $62 million (just under 50 percent). Had the payout ratio for Katrina been the same as for Andrew, insurers would have paid out $80 billion, or $18 billion more than they did. The bottom line in these comparisons is that, if insurers had not reduced policyholder coverage and increased rates after Hurricane Andrew, they would have paid out almost 30 percent more to them How is it possible that the property-casualty industry’s surplus would sharply increase as the number and severity of catastrophic weather events also increases? The primary reason is that the insurers have ―mastered‖ hurricanes by shifting the lion’s share of the risk and costs to consumers and taxpayers. In other words, property-casualty insurers have paradoxically emerged as masters of risk avoidance, rather than continuing their historic role of risk taking.
HOW INSURERS REDUCED THEIR HURRICANE LOSSES AND SHIFTED RISK TO CONSUMERS AND TAXPAYERS First, insurers have made intelligent use of reinsurance, securitization and other risk spreading techniques.6 Some insurers now spread risk by issuing securities that couple the threat of a catastrophic event with the purchase of construction stocks that would likely increase in value if a catastrophic event occurs and the demand for construction increases. The use of this kind of creative approach to diversify risk is wise. Second, after Hurricane Andrew, insurers changed ratemaking techniques by using computer models to project either 1,000 or 10,000 years of weather experience. While this caused huge price increases to consumers at the time, consumer leaders supported this change because insurers appeared to be genuinely surprised by the level of damage caused by Hurricane Andrew and promised that the models would bring long-term stability to prices. The model contained projections of periods of intense activity and very large hurricanes, as well as periods of little or no activity, and based rates on these estimates. However, Risk Management Solutions (RMS) and the other risk modeling companies have recently stopped using this scientific method to project storms over a 1,000 or 10,000-year period and are now using one to five-year projections. This has caused at least a 40 percent jump in loss projections in Florida and the Gulf Coast and a 25 percent jump in the Northeast. This move reneges on promises of pricing stability made by insurers in the mid-1990s and has led to rates that are excessive. Insurance rates on the coasts have soared for property risks, homes and businesses in the last few years. Third, insurers have sharply hollowed out the catastrophe coverage offered to consumers in recent years by placing a number of new requirements on policyholders and limits on coverage in policies: Deductibles of 2 to 5 percent have been imposed with little fanfare or notice. This reduction in coverage was accompanied in many cases by large rate increases. Caps on replacement costs and other limits on needed coverage. State Farm, for instance, caps payments for increased rebuilding costs at 20 percent. Other insurers allow no increased payments at all. A consumer who buys a $100,000 policy would receive only $100,000 to rebuild from some insurers, and $120,000 from State Farm, even if the cost of repairs skyrockets after a storm due to increased demand for materials and labor. Costs can also increase when homeowners are required to make special repairs to comply with building codes that were enacted after a home was first constructed. For example, many municipalities require such code upgrades to comply with the National Flood Insurance Program if a home is more than 50 percent damaged by a flood.
This report is focused on the primary insurance market, not the reinsurance market. The worldwide reinsurance market has had rather stable catastrophe prices since 2002, with ―rates on line‖ – defined by the reinsurer Guy Carpenter as ―Premium divided by indemnity (claims paid). A British term for the rate which, when multiplied by the indemnity, would produce the premium.‖ in a tight range since 2002. The highest rate on line observed in the data was in 1993 as Andrew severely impacted pricing. Today prices are stable because the catastrophe reinsurance sector ―was overcapitalized by more than US $20 billion, or 12 percent at the beginning of 2010.‖ This led to share buy-backs by many reinsurers. (Material in this footnote based on ―World Catastrophe Reinsurance Market,‖ Guy Carpenter, September 2010.)
Reimbursement for costs incurred to comply with building codes is now excluded from many homeowners’ insurance policies. Coverage for mold mitigation is also now excluded from most policies. Given the surge in demand for home building and repair that occurs in the wake of a hurricane, and corresponding increases in prices, and new coverage exclusions, these changes significantly shift risk and costs to consumers. ―Anti-concurrent-causation‖ clauses. This is the most draconian reduction in coverage that insurers have attempted to impose in recent years. It removes all coverage for wind damage if another, non-covered event (usually a flood) also occurs, regardless of the timing of the events. Under this anti-consumer measure, if a hurricane of 125-miles-perhour rips a house apart but hours later a storm surge floods the property, the consumer would receive no reimbursement for wind losses incurred. Given the cutbacks in coverage that have occurred in coastal areas, there is a serious question as to whether this diminished coverage is worth the higher rates that many consumers must pay. However, most consumers have no option but to purchase such coverage as it is required by lenders or the law or both. Demand for insurance is relatively inelastic. Insurers have claimed that they are facing higher risks because of a sharp increase in the number of people and amount of construction in areas of the country vulnerable to earthquake and hurricane disasters. This claim was investigated in 2006 by the Los Angeles Times investigative reporter Peter Gosselin, who wrote that: …Key statistics don’t support the argument….Census figures…show that the population of coastal and earthquake counties grew at an annual average rate of 1.56 percent between 1980 and last year. But they show that the U.S population grew at a reasonably close pace of 1.24 percent. Gosselin interviewed Judith T. Kildow, director of the government-funded National Ocean Economics Program at California State University at Monterey, who said, ―You simply cannot make the case from the numbers that America’s coastal counties have grown at a disproportionately faster rate than the country as a whole over the last 25 years.‖7 Fourth, insurers have also shifted risk, sometimes onto taxpayers. Taxpayers are exposed by the high deductibles, anti-concurrent causation and other limits on coverage as disaster relief will fill in what insurers used to cover. Taxpayers might also be called upon to subsidize state-run insurers-of-last resort, which were sharply populated by insurers non-renewing tens of thousands of homeowner and business properties. Allstate, the leading exemplar after Hurricane Andrew, emerged once again as the company that was most aggressive in refusing to renew homeowner’s policies in the wake of Hurricane Katrina. After Hurricane Andrew, Allstate threatened to non-renew 300,000 South Floridians, leading the state of Florida to place a moratorium on such precipitous actions. After Hurricane Katrina, Allstate non-renewed thousands of homeowners, even many on Long Island, New York and Cape Cod, Massachusetts. Allstate has also announced that it will no longer offer new homeowner’s policies in many states, from Connecticut to Delaware, and has refused to
―The New Deal – Insurers Learn to Pinpoint Risks – and Avoid Them,‖ Peter Gosselin, Los Angeles Times, November 28, 2006.
write new business in large portions of other states, such as Maryland and Virginia. Other insurers have also cut back coverage on the nation’s coasts (See Addendum B, for more information). Insurers have become quite adept at convincing government to use tax dollars to help them avoid risk. Consider the federal Terrorism Risk Insurance Act (TRIA), the California Earthquake Authority, Citizen’s Insurance in Florida, and wind ―pools‖ in a number of other states. The state pools have become the largest writers of insurance in some states8. Such an arrangement allows insurers to ―cherry-pick‖ these states, keeping the safest risks for themselves and shifting the highest risks onto the taxpayers of the state, thereby socializing high-risk, potentially unprofitable policies and privatizing the low-risk, profitable business. This adverse result for policyholders and taxpayers is hardly surprising. It is akin to ―solving‖ the health insurance crisis by requiring states to cover sick or terminally ill patients, while the private sector writes coverage for young and healthy consumers. Allstate has also led efforts at the federal level that failed to create a taxpayer-backed program modeled on TRIA to reinsure the private market against the perils of wind and other weather damage. INSURERS COULD EASILY HANDLE CATASTROPHE RISK THEY ARE AVOIDING BECAUSE THEY ARE SIGNIFICANTLY OVERCAPITALIZED In determining whether the property-casualty insurance industry is adequately capitalized, one must first examine the losses incurred for major catastrophe or terrorism events. According to the Insurance Information Institute, the top ten insured loss disasters for property were: PRE-TAX POST TAX EVENT9 2010 DOLLAR LOSS 1. Hurricane Katrina, August 2005 2. World Trade Center, Pentagon terrorist attacks, September 2001 3. Hurricane Andrew, August 1992 4. Northridge, California earthquake, January 1994 5. Hurricane Ike, September 2008 6. Hurricane Wilma, October 2005 7. Hurricane Charley, August 2004 8. Hurricane Ivan, September 2004 9. Hurricane Hugo, September 1989 10. Hurricane Rita, September 2005 $45.5 billion $29.6 22.9 14.9 22.4 14.6 17.3 11.2 12.7 8.3 11.4 7.4 8.5 5.5 8.1 5.3 6.7 4.4 6.2 4.0
Source: Insurance Services Office (ISO); Insurance Information Institute (See http://www.iii.org/facts_statistics/catastrophes-us.html). (Ranked on constant dollar cost to insurers)
According to PIPSO – The Property Insurance Plans Service Office, the Florida FAIR Plan had 1.5 million policies, of which over 285,000 were high-risk coastal properties on May 2011. The Texas wind pool had 247,972 residential and 17,998 commercial policies in 2010. In 2010, Alabama’s pool had 18,800 policies (mere than double the 7,800 of 2007z), Mississippi’s Beach/Windstorm Plan had 46,546 policies and Georgia had 26,340 policies in its Pool. 9 The catastrophes were ranked by III based on size of loss in 2005 dollars, which we do not display here. What is displayed is the actual dollars in the year of the event. We calculate the post-tax figure by deducting the corporate tax rate of 35 percent.
Considering that property-casualty insurers now have surplus in excess of $580 billion,10 catastrophes of this size are very easy to manage. Terrorism risk is an interesting case study. While insurers are rightly concerned about a huge event, such as a nuclear, chemical or biological attack, the actual terrorism events that have occurred so far have been easily managed by private industry. There were hundreds of terrorism events in America in the 20 years leading up to the September 11th attacks. In spite of this fact, insurers did not even bother to charge a separate price for terrorism coverage in their rating structures. September 11th changed this practice, but even that attack was a ―small‖ insured event compared to the industry’s mammoth capital and surplus, which has grown significantly since 2001. Yet, insurers convinced the federal government to provide free reinsurance that CFA estimates has represented about a ten billion taxpayer subsidy to date. Historically, the prime test for the solidity of the property-casualty insurance industry has been the ratio of net premiums written (NPW) to surplus, discussed above. Regulators became concerned about the financial soundness of an insurer if its ratio exceeded 3 to 1. The so-called ―Kenney Rule,‖ named after financial writer Roger Kenney, held that a safe insurer should not exceed about a 2 to 1 ratio. This guideline was introduced in the 1960s and served as the standard that insurers and regulators followed for many decades. More recently, analysts have recommended lowering the acceptable ratio to about 1.5 to 1, in recognition of some more extreme risks that insurers now face, such as catastrophic hurricanes and terrorist attacks. Net premium written to surplus ratios for almost thirty years are as follows: NPW/SURP YEAR 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
1.80 1.81 2.14 2.20 1.84 1.63 2.00 2.81 2.50 2.40 2.48 2.34 2.14 1.85 1.83 1.73 1.65 1.86 1.91 1.88 1.87 1.71
As of December 31, 2010, Bests Aggregates and Averages, 2011 Edition, page 366.
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Source: Best's Aggregates and Averages, 1988-2011
1.56 1.58 1.40 1.40 1.33 1.30 1.13 1.05 0.90 0.84 0.85 0.93 1.10 1.28 1.16 1.09 1.00 0.90 0.84 0.94 0.80 0.74
Property-casualty insurers have not exceeded the recommended 1.5 to 1 ratio of NPW to surplus in almost twenty-five years. The sharp downward trend in this key leverage ratio is very clear, demonstrating that the industry is now significantly overcapitalized. Here is a graphic display of these data:
3 2.5 2 1.5 1 0.5 0 YEAR 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
Consider this startling fact: Even if all of the top ten catastrophic events, including the September 11, 2001 attack, the Northridge Earthquake, and the top eight hurricanes, had 8
occurred in the last year and had been paid for last week (a total of $162 billion in 2010 dollars after tax11), the property-casualty industry surplus would still be at $418 billion and the leverage ratio would still be at an ultra safe ratio 1.0.12 WHO PAYS WHEN INSURERS DO NOT? Consumers Data indicates that Hurricane Katrina cost $125.0 billion, of which $62.2 billion (just under 50 percent) was paid by insurance. Hurricane Andrew cost $26.5 billion, of which $17.0 billion (64 percent) was paid by insurance.13 To show the difference in coverage now that the policies have been hollowed out, consider a hypothetical $100,000 home that incurred different levels of wind damage under the Hurricane Andrew compared to Hurricane Katrina. Assume the home had a $500 deductible under Andrew and a 5 percent deductible under Katrina. Benefit after Deductible Andrew Katrina Katrina as a % of Andrew
$10,000 $ 9,500 $ 5,000 53.6% (Consumer pays $500 in Andrew; $5,000 in Katrina) $50,000 $49,500 $45,000 90.9% (Consumer pays $500 in Andrew; $5,000 in Katrina) Assume further that additional work must be done when it is reconstructed to bring it up to code. If, for instance, the home required $1,000 of electrical work, the policyholder would be paid an additional $1,000 for Hurricane Andrew in both circumstances. However, under Katrina, there would the policyholder would receive no additional payment under the policy for mandated code work. Benefit after Deductible Katrina as a % of Andrew Damage Andrew Katrina $10,000 $ 9,500 + $1,000 $ 5,000 (Consumer pays $500 in Andrew; $6,000 in Katrina) $50,000 $49,500 + $1,000 $45,000 (Consumer pays $500 in Andrew; $6,000 in Katrina) 47.6%
If the home was in a flood plain and not elevated, damages that totaled 50 percent of the home’s value would trigger a ―non-conforming use‖ under the National Flood Insurance
From Insurance Information Institute at http://www.iii.org/facts_statistics/catastrophes-us.html. $430 million in premium divided by ($580 million in surplus less $162 million in assumed after-tax loss) Data on Net Premiums Written and Surplus for all insurers is from A. M. Best Aggregates and Averages, 2011 Edition, Page 369. 13 See http://www.iii.org/facts_statistics/hurricanes.html.
Program and the home would have to be upgraded to withstand a ―100-year‖ flood. If such an improvement costs $10,000, the damage situation with $50,000 in losses would be: Benefit after Deductible Andrew Katrina as a % of Andrew Katrina 24.4%
$10,000 $ 9,500 + $1,000 + $10,000 $ 5,000 (Consumer pays $500 in Andrew; $6,000 in Katrina)) $50,000 $49,500 + $1,000 + $10,000 $45,000 (Consumer pays $500 in Andrew; $16,000 in Katrina)
It could get even worse if the home is destroyed and a demand surge of 50 percent raised the rebuilding cost for the $100,000 home to $150,000. If there was any additional flood damage to the home, even minor damage, (such as $5,000) the insurer might invoke the anticoncurrent causation clause of the policy and pay nothing. If the damage is caused only by wind, but the severe damage in the area causes rebuilding costs to rise by 50 percent, the benefit situation would be: ANDREW: $150,00014 less $5,000 for flood damage plus $10,000 flood elevation less $500 deductible = $154,500. Consumer pays $5,500. KATRINA: Insurance pays nothing. Consumer pays $160,000. It is very clear that much of the cost that used to be paid by private insurers has been shifted to consumers, ranging from a small amount to 100 percent of what used to be paid. Much of this will be shifted again, to taxpayers, in the form of increased disaster relief payouts as discussed now. Taxpayers Taxpayers are also bearing more risk. The National Flood Insurance Program is almost $20 billion in debt because the program is poorly administered by the Federal Emergency Management Agency (FEMA) and is poorly designed. FEMA has allowed flood risk maps to become antiquated, which has resulted in inadequate premiums that encourage unwise construction. The ―Write Your Own‖ program which requires taxpayers to shoulder all financial risk, but allows insurers to service the NFIP, has also allowed insurers to collect excessive fees. The program will surely go deeper in debt in coming years due to these among other hidden subsidies, such as ―grandfathering‖ in low, inadequate rates when new maps are issued. Congress should require the private sector to take a small, but growing, percentage of the risk over time. This will reduce taxpayer exposure in two ways: (1) The private sector will pay for flood losses incurred on their own accounts and (2) once the private sector has financial responsibility for some flood losses, it will police against hidden subsidies, such as FEMA’s unauthorized (by Congress) grandfathering of low rates on supposedly actuarially-rated homes when a new map raises flood elevations.
Some companies, such as State Farm, would pay an additional 20% or $30,000 more in this example.
The subsidy to taxpayers under TRIA has amounted to roughly $10 billion. Although there is no need for the federal government to back terrorism risk (except for exotic risks like nuclear and biological attacks), insurers are enjoying what amounts to free reinsurance, making TRIA another example of a wasteful corporate subsidy. When private insurance payouts decline during catastrophic weather events, it stands to reason that government disaster relief costs will increase. Hurricane Andrew generated public disaster relief payments of $7 billion in 2009 dollars, whereas Hurricane Katrina generated $51 billion.15 Insurance payments were twice as high for Katrina as for Andrew but disaster relief payments were over seven times higher. In 2008, with Hurricanes Gustav and Ike, among others, federal disaster relief was $13 billion (2009 dollars). When storms like these eclipse Andrew in the amount of government disaster relief that is paid out, it is hard not to conclude that a major cause of this increased taxpayer burden is the reduction in risk carried by insurers. The total cost of disaster relief from 1990 to 1999 was $40 billion; from 2000 to 2010 it was more than double, at $94 billion (all in 2009 dollars). Because private insurers have fled America’s coasts, many homes are insured through state pools. In Florida, Florida Citizens’ Property Insurance Corporation (CPIC), the state pool, covers 1.5 million homes. If a storm or series of storms depletes Citizens’ rather healthy reserves, assessments may be placed on other property-casualty companies in the state to address the shortfall. The state has the authority, since assessments are limited, to finance loss payments via tax-exempt bonds. State taxpayers could be at some risk if Citizens’ or the CAT Fund (also backed by the state) ever ran out of money, even though this possibility has become less likely as the state has built up reserves. In California, earthquake risk is mostly written through the California Earthquake Authority, an entity similar to Florida Citizens, where state taxpayers may be exposed in extreme events. RECOMMENDATIONS There are many conclusions to be drawn from the reduced losses that insurers have experienced in recent years. The prime conclusion is that the insurance industry has moved from its historic role as a calculated risk-taker to one of a risk-avoider, exposing consumers and taxpayers to much higher costs. Not only have insurers insulated themselves from their historic share of hurricane risk, they have made no serious effort to write flood risk and terrorism risk, which are entirely backed by federal taxpayers. Although insurers have become adept at shifting the cost of catastrophe losses to others, they still use catastrophic weather events to advocate for measures that would shift risk even more, such as higher rates, or putting more policyholders in pools or created taxpayer-supported entities. Thus, many consumers exposed to catastrophe weather risk are also vulnerable to insurer attempts to unjustifiably increase rates or hollow out coverage. Recommendations for the States CFA recommends that the states carefully examine national data on limited catastrophe losses and excessive surplus before approving any insurer requested rate increases. State
Database maintained by Congressional Research Service based upon US Budget documents and appropriation statutes, Table 1 of ―Disaster Relief Funding and Emergency Supplemental Appropriations, CRS, May 24, 2010.
insurance commissioners should be on guard against unwarranted attempts by insurers to use catastrophe losses as part of their rationale for jacking up rates. We recommend that states carefully review the reasons why insurers are dumping risks into state pools and to take action to stop insurers from unjustifiably refusing to cover qualified homeowners. It is unnecessary for any more dumping to occur since insurers have now twice purged their portfolios of risk, once after Hurricane Andrew and again after Hurricane Katrina. States should also look at the high prices being charged to homeowners in their states in light of the fact that, in the aftermath of Andrew, insurers made major adjustments to pricing, dumped risk, reduced coverages and significantly reduced their hurricane risk exposure. Repeating these adjustments in the wake of 2004/5 storms was really more about gouging than correction. Rates requested by insurers in non-competitive markets after hurricanes can easily be excessive, violating sound actuarial principles. The coastal states must revisit hurricane pricing in recognition of the fact that the industry has mastered hurricanes on a national basis as evidenced by the almost negligible impact of Hurricane Katrina on their national results, a vast change compared to results during Hurricane Andrew, a smaller event. States should ban the use of nonscientific pricing models, such as short-term catastrophe models. States should ban use of anti-concurrent causation clauses and any other attempt by insurers to build a ―trap-door‖ hidden in the policy, through which coverage can unexpectedly fall when policyholders most need help. States should not allow hurricane deductibles to apply unless a storm is classified as a hurricane throughout its journey through the state, from entry to exit. It is impossible to tell where a hurricane exactly becomes a tropical storm within a state so this ambiguity must be decided to the benefit of the consumers who have bought the coverage, not insurers who are compensated for being risk-takers. States should adopt the California approach to consumer participation in regulatory proceedings, where consumers can receive reimbursement from the filing insurer to hire experts (like actuaries, lawyers and economists) if they make a ―substantial contribution‖ to a case. They receive no compensation if they do not make a substantial contribution, so consumer groups in California study filings prior to risking an intervention very carefully. Costs paid by the insurers for such intervention would be allowed to be included as part of the rate filing. States should make sure that they have all the data they need to monitor the home insurance market, including data by census track on who is writing and where, non-renewal patterns, etc. This will allow regulators to make informed decisions about whether markets in their states are truly competitive. Coastal states should join together to form an interstate compact to deal with common issues stemming from their shared hurricane risk. A pool of states with common policies could allow states to spread risk and lower costs by developing common pools and provide consumers and insurers with consistent requirements. A common approach would also better position states – especially small ones – to resist coercive efforts by insurers to weaken regulatory protections for consumers. For example, after the hurricanes of 2004/5, several smaller states (AL, MS and LA) were pressured by insurers with threats of withdrawal to take actions that would harm consumers in those states. 12
One action a consortium of coastal states could take is to create a regulatory model for calculating hurricane risk to test the reliability of insurer-proposed catastrophe premiums in rates. Using the model, states could create a stand-by reinsurance mechanism that would sell reinsurance to insurers at 50 percent more than actuarial rates developed by the model, which would keep premiums in check during the non-competitive phase of the insurer cycle or after hurricane events when reinsurers often gouge. When private reinsurance is reasonably priced at or near the actuarial level, the state back-up would not kick in. Florida successfully did sell reinsurance to the industry after the 2004/5 storms and now has, through premium accumulation and bonding, no real risk for its reinsurance. At the same time, policyholders saved about 15 percent of premiums because Florida requires insurers to use their reinsurance (or at least adopt the cost of that state reinsurance or less in home insurance ratemaking). This system temporarily replaced private reinsurance in Florida, which was priced at four to five times the actuarial rate in the non-competitive reinsurance market after 2004 and 2005 storms. States should also develop model language that would be required in every minimum insurance policy sold in the region. Among other things, this language should remove the anticoncurrent causation clauses from use and clarify exceptions and exclusions in coverage. Coverage above the minimum would be allowed to be sold to consumers with pricing for such enhancements made clear to the policyholder. Recommendations for the Federal Government The fact that insurers do not take financial risk for either flood or terrorism insurance is a huge policy error. With the NFIP, it tempts unscrupulous insurers to illegitimately shift wind risk to the flood program. With both programs, taxpayers are required to pick up huge risks that private insurers are more than capable of identifying and backing. Taxpayers deserve to have at least some of this risk removed from them, particularly at this time of economic stress, and a search for ways to cut federal spending. We recommend that Congress limit the exposure of taxpayers to terrorism risk to only extreme events such as nuclear, chemical or biological events exceeding a 100 billion threshold. TRIA should be amended to only cover losses caused by nuclear, biological or chemical attacks that exceed $100 billion. We recommend that the National Flood Insurance Program bills currently under consideration be amended to require a study on how to involve the private sector in sharing the risk from the first dollar of loss, perhaps starting with a low, but increasing, percentage of the risk for insurers wishing to participate in the NFIP as ―Write Your Own‖ companies.16 FEMA should also lower the excessive WYO servicing fees that create a windfall for the WYO companies at taxpayer expense. Recommendations for private reinsurers to cover losses only above the federal insurance coverage that is offered should be rejected, as they will only add the cost of reinsurers’ overhead and profit to the program over time. It does not make sense to have the relatively tiny reinsurance industry backing up the federal government, at the same time as insurers are pushing for a federal backstop for their wind exposure. Logically, a smaller entity should not be backing up a bigger one, which argues for a private NFIP role at the low end of flood loss spectrum, not at the high end. If the program were privately reinsured, it would most
See June 23, 2011 Testimony given by Travis Plunkett to the US Senate Banking Committee on the ―Authorization of the National Flood Insurance Program‖ for detailed comments on CFA’s recommendations.
likely add the unnecessary cost of the overhead and profit the reinsurers receive to the cost of the program, requiring more taxpayer expenditures over time. Further, if significant events occur, reinsurers are likely to either retrench or severely raise prices, just when reinsurance is most needed, as they do with wind coverage. Data on home insurance in a format similar to what is required of banks under the Home Mortgage Disclosure Act (HMDA) should be collected by the Federal Insurance Office (FIO) and made available to the states and to the public. This would allow detailed analysis of why certain markets are stressed, which insurers are doing their best to serve markets in stressed areas and which are causing problems. This would enable policyholders to craft solutions based on solid statistical evidence. It would also allow analysis of markets to see if low- and moderateincome areas are being properly served. We recommend that the federal government assist the states in forming an interstate compact to regulate hurricane insurance by authorizing such a combined effort and by taking action to assist the states in several ways, including: Offering the expertise of the federal government (entities like FEMA, NOAA, etc.) to the group of coastal states. These experts could help develop the regulatory hurricane model for states to use in regulating insurance and in developing stand-by reinsurance pricing. Offering bridge loans at low-interest when stand-by reinsurance is used, if such use suffers losses due to timing risk (such as a large storm in the early years of the development of reserves.) These loans would be required to be fully repaid over reasonable time periods.
ADDENDUM A DATA UNDERLYING THE CHARTS FOUND IN THE BODY OF THIS REPORT
COLUMN 1 YEAR COLUMN 2 Total P/C Op Inc/Prem 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 4.6% 3.7% 2.3% 5.0% 9.3% 9.9% 7.6% 2.2% -0.7% 4.2% 10.1% 10.9% 9.2% 8.2% 7.1% 4.5% 2.5% -3.4% -4.2% 3.2% 7.3% 8.0% 5.0% 5.2% 6.2% -1.1% 6.2% 4.8% 7.7% 7.9% 13.1% 8.3% 5.3% 3.5% -4.0% 2.4% 8.9% 10.1% 10.5% 19.8% 16.9% 7.2% COL 3 COLUMN 4 COL 5
YEAR P/C L/R 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 66.8% 69.4% 70.7% 77.1% 77.0% 70.2% 66.6% 66.4% 69.2% 69.4% 68.5% 74.4% 66.7% 68.1% 65.7% 65.4% 60.3% 63.1% 65.2% 68.3% 75.3% 68.8% 62.2% 60.2% 61.6% 53.5% 56.0% 62.3% HO L/R 63.0% 64.5% 63.7% 66.4% 70.7% 61.9% 56.0% 59.1% 70.9% 70.8% 74.8% 114.1% 70.9% 74.9% 69.6% 78.6% 57.3% 65.3% 65.0% 69.0% 79.4% 68.5% 59.0% 56.5% 60.4% 50.0% 55.6% 74.4%
4 FLORIDA HURRICANES HURRICANE KATRINA
Source: All data from Best's Aggregates and Averages, various years. Columns 1 and 2 were used to create the chart showing the Property-casualty Insurance Industry’s Economic Cycle over more than four decades. Columns 3, 4 and 5 were used to create the chart showing the loss ratios for both the overall Property-casualty Insurance Industry and for homeowners insurance over a thirty-year period.
Addendum B: Reprinted from The Los Angeles Times, November 28, 2006
Insurance company cutbacks have left more than 1 million coastal residents scrambling to land new insurers or learning to live with weakened policies. As insurers retreat, states and homeowners are left to bear the biggest risks. Massachusetts During the last two years, six insurers have stopped selling or renewing policies along the coast, especially on Cape Cod, leaving 45,000 homeowners to look for coverage elsewhere. Most have turned to the state-created insurer of last resort. The Massachusetts FAIR Plan, now the state's largest homeowners insurer, recently received permission to raise rates 12.4 percent. Connecticut Atty. Gen. Richard Blumenthal has subpoenaed nine insurance companies to explain why they are requiring thousands of policyholders whose houses are near any water —coast, river or lake —to install storm shutters within 45 days or have their coverage cut or canceled. New York Allstate has refused to renew 30,000 policies in New York City and Long Island, and suggested it may make further cuts. Other insurers, including Nationwide and MetLife, have raised to as much as 5 percent of a home's value the amount policyholders must pay before insurance kicks in, or say they will write no new policies in coastal areas. South Carolina Agents say most insurers have stopped selling hurricane coverage along the coast. Those that still do have raised their rates by as much as 100 percent. The state-created fallback insurer is expected to more than double its business from 21,000 policies last year to more than 50,000. Florida Allstate has offloaded 120,000 homeowners to a start-up insurer and has said it will drop more as policies come up for renewal. State-created Citizens Property, now the state's largest homeowners insurer with 1.2 million policies, was forced to use tax dollars and issue bonds to plug a $1.6- billion financial hole due to hurricane claims. The second-largest, Poe Financial Group, went bankrupt this summer, leaving 300,000 to find coverage elsewhere. The state also has separate funds to sell insurers below-market reinsurance and cover businesses. Controversy over insurance was a major issue in this fall's election campaign, causing fissures in the dominant GOP.
Louisiana The state's largest residential insurer, State Farm, will no longer offer wind and hail coverage as part of homeowners policies in southern Louisiana. In areas where it still covers these dangers, it 17
will require homeowners to pay up to 5 percent of losses themselves before insurance kicks in. In a move state regulators call illegal and are fighting, Allstate is seeking to transfer wind and hail coverage for 30,000 of its existing customers to the state created Citizens Insurance. Texas Allstate and five smaller insurers have canceled hurricane coverage for about 100,000 homeowners and have said they will write no new policies in coastal areas. Texas' largest insurer, State Farm, is seeking to raise its rates by more than 50 percent along the coast and 20 percent statewide. California The state has bucked the trend toward higher homeowners insurance rates with three major insurers, State Farm, Hartford and USAA, seeking rate reductions of 11 percent to 22 percent. Regulators have begun to question whether insurers are making excessive profits after finding that major companies spent only 41 cents of every premium dollar paying claims and related expenses. Alone among major firms, Allstate is seeking a 12.2 percent rate hike. Washington Allstate has dropped earthquake coverage for about 40,000 customers and will have its agents offer the quake insurance of another company when selling homeowners policies in the state. Nationally, the company has canceled quake coverage for more than 400,000. Sources: Risk Management Solutions (map); interviews with state insurance regulators
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.